DroppGroup
AMER
AI/ML, SaaS
AWS Sagemaker HyperPod, AWS Sagemaker Endpoints, AWS EC2, AWS ELB
DroppGroup leads in integrating AI and blockchain technology with patented innovations like the aMiGO Foundational AI model. Trusted by industry giants such as Saudi Aramco, Cisco, NVIDIA and the Saudi Government, DroppGroup transforms the 2D Internet into 3D immersive experiences and 3D print-ready outputs in just 5 seconds without any coding. Partnering with Avahi and utilizing AWS services, DroppGroup improved precision and inference times for 3D model generation, resulting in a scalable, efficient solution that enhances user experience and operational efficiency.
DroppGroup Challenge: Improving Precision and Reducing Inference Times
DroppGroup technology challenge was achieving high precision and reducing the 30-second inference times required for generating high poly 3D models from as little as one single 2D image. The models struggled to capture intricate details accurately, leading to lower-quality 3D models and system timeouts. These issues also hindered scalability, as the system needed to manage high volumes of concurrent computations.
To address these challenges, DroppGroup turned to AWS, focusing on key services such as SageMaker Endpoints and SageMaker HyperPods. AWS SageMaker Endpoints provided a robust platform for deploying machine learning models that could handle real-time, scalable inferences, significantly reducing latency. SageMaker HyperPods enabled distributed and accelerated training of models across multiple compute instances, which was crucial for efficiently managing large-scale image data. Additionally, HyperPods’ ability to automatically detect and replace faulty nodes with healthy ones ensured system reliability.
Avahi played a critical role in migrating DroppGroup infrastructure to AWS and optimizing the overall system performance. Avahi conducted a comprehensive readiness assessment to identify potential bottlenecks and optimize the deployment process. They also implemented AWS cost management tools, ensuring that all compute resources were properly tagged and monitored. Avahi’s expertise helped DroppGroup fine-tune their models and infrastructure, significantly enhancing precision and reducing inference times.
What Changes AWS and Avahi Brought for DroppGroup
The combined efforts of AWS and Avahi brought remarkable improvements to DroppGroup application. By utilizing p5 instances on AWS SageMaker Endpoints, DroppGroup reduced latency from 30 seconds to just 5 seconds. The scalability offered by AWS services, including AWS EC2 and AWS ELB, enabled DroppGroup to rapidly expand its user base from 0 to over 100,000 in just a few weeks. This scalability also streamlined the onboarding process, reducing the time required from weeks to days, thus enhancing operational efficiency and user satisfaction. Moreover, DroppGroup benefited from the robust security and reliability provided by AWS, ensuring the integrity of user data and models within its Virtual Private Cloud (VPC).
Founder, Bravo Foxtrot
